from __future__ import annotations from typing import Iterable import gradio as gr from gradio.themes.base import Base from gradio.themes.utils import colors, fonts, sizes from llama_cpp import Llama from huggingface_hub import hf_hub_download hf_hub_download(repo_id="Pi3141/alpaca-lora-7B-ggml", filename="ggml-model-q4_1.bin", local_dir=".") llm = Llama(model_path="./ggml-model-q4_1.bin") ins = '''Below is an instruction that describes a task. Write a response that appropriately completes the request. ### Instruction: {} ### Response: ''' ins_inp = '''Below is an instruction that describes a task. Write a response that appropriately completes the request. ### Instruction: {} ### Input: {} ### Response: ''' theme = gr.themes.Monochrome( primary_hue="indigo", secondary_hue="blue", neutral_hue="slate", radius_size=gr.themes.sizes.radius_sm, font=[gr.themes.GoogleFont("Open Sans"), "ui-sans-serif", "system-ui", "sans-serif"], ) def generate( instruction, input=None, temperature=0.1, top_p=0.75, top_k=40): result = "" if input: instruction = ins_inp.format(instruction, input) else: instruction = ins.format(instruction) for x in llm(instruction, stop=['### Instruction:', '### End'], stream=True, temperature=temperature, top_p=top_p, top_k=top_k): result += x['choices'][0]['text'] yield result examples = [ "Instead of making a peanut butter and jelly sandwich, what else could I combine peanut butter with in a sandwich? Give five ideas", "How do I make a campfire?", "Explain to me the difference between nuclear fission and fusion.", "Write an ad for sale Nikon D750." ] def process_example(args): for x in generate(args): pass return x css = ".generating {visibility: hidden}" # Based on the gradio theming guide and borrowed from https://huggingface.co/spaces/shivi/dolly-v2-demo class SeafoamCustom(Base): def __init__( self, *, primary_hue: colors.Color | str = colors.emerald, secondary_hue: colors.Color | str = colors.blue, neutral_hue: colors.Color | str = colors.blue, spacing_size: sizes.Size | str = sizes.spacing_md, radius_size: sizes.Size | str = sizes.radius_md, font: fonts.Font | str | Iterable[fonts.Font | str] = ( fonts.GoogleFont("Quicksand"), "ui-sans-serif", "sans-serif", ), font_mono: fonts.Font | str | Iterable[fonts.Font | str] = ( fonts.GoogleFont("IBM Plex Mono"), "ui-monospace", "monospace", ), ): super().__init__( primary_hue=primary_hue, secondary_hue=secondary_hue, neutral_hue=neutral_hue, spacing_size=spacing_size, radius_size=radius_size, font=font, font_mono=font_mono, ) super().set( button_primary_background_fill="linear-gradient(90deg, *primary_300, *secondary_400)", button_primary_background_fill_hover="linear-gradient(90deg, *primary_200, *secondary_300)", button_primary_text_color="white", button_primary_background_fill_dark="linear-gradient(90deg, *primary_600, *secondary_800)", block_shadow="*shadow_drop_lg", button_shadow="*shadow_drop_lg", input_background_fill="zinc", input_border_color="*secondary_300", input_shadow="*shadow_drop", input_shadow_focus="*shadow_drop_lg", ) seafoam = SeafoamCustom() with gr.Blocks(theme=seafoam, analytics_enabled=False, css=css) as demo: with gr.Column(): gr.Markdown( """ ## Alpaca-LoRa 7b quantized 4bit (q4_1) Type in the box below and click the button to generate answers to your most pressing questions! """ ) with gr.Row(): with gr.Column(scale=3): instruction = gr.Textbox(lines=2, placeholder="Tell me more about alpacas.", label="Instruction", elem_id="q-input") input = gr.components.Textbox(lines=2, label="Input", placeholder="none") temperature = gr.components.Slider(minimum=0, maximum=1, value=0.1, label="Temperature") top_p = gr.components.Slider(minimum=0, maximum=1, value=0.75, label="Top p") top_k = gr.components.Slider(minimum=0, maximum=100, step=1, value=40, label="Top k") with gr.Box(): gr.Markdown("**Output**") output = gr.Markdown(elem_id="q-output") submit = gr.Button("Generate", variant="primary") gr.Examples( examples=examples, inputs=[instruction], cache_examples=False, fn=process_example, outputs=[output], ) submit.click(generate, inputs=[instruction, input, temperature, top_p, top_k], outputs=[output]) instruction.submit(generate, inputs=[instruction], outputs=[output]) demo.queue(concurrency_count=1).launch(debug=True)